What is Network Observatory

What it is

Network Observatory is the real-time monitoring module of CreaRack Pro. It watches every device in your infrastructure that has an IP address — switches, routers, servers, access points, UPS units, signage players — and tells you what is up, what is slow and what is failing before your users notice.

Each monitored device gets three independent check types. Enable any combination:

• Ping (ICMP) — availability, latency and packet loss. The baseline check.
• SNMP — bandwidth per interface (v2c or v3), interface errors and discards, plus CPU, memory and temperature on devices with known vendor OIDs.
• HTTP — health checks for web services: status code, response time and SSL certificate validity. See [[crearack—monitoring—ping-http]].

On top of the raw checks, Observatory adds configurable alerts, an AI analysis layer (CNS), an ITSM tab for incident management, and a Network Tutor chat for networking questions.

Two monitoring modes

The Sentinel Mode panel in the sidebar controls who runs the checks:

Click Change in the panel to switch. In Sentinel mode the Agent runs the checks 24/7 and pushes the results to the cloud. In both modes, how often each device is checked is up to you: every device has its own Polling cadence (15 s to 5 min, default 1 min), set in its card editor — see [[crearack—network—device-page]]. Agent status and management live in the Agent Fleet widget — see [[crearack—monitoring—fleet-manager]].

The interface

• Sidebar (left) — your device list with a Devices / Manual source toggle, free-text search and filters by type, status, group and rack. Below it: the Sentinel Mode panel, Chart Defaults, alert severity cards (Good / Info / Critical) and the Network Tutor button. Drag the edge to resize (280-560 px).
• Workspace (right) — fixed tabs Overview (fleet dashboard), CNS (AI insights) and ITSM (incidents), plus one tab per device you open. Open tabs are saved and restored on reload.

The sidebar also has a Deep Discovery button that probes all SNMP targets for vendor details and extra metrics — see [[crearack—monitoring—deep-discovery]].

How to start monitoring a device

1. Devices managed in CreaRack (racks, Auto-Provision) appear under Devices automatically. For anything else, click Add and enter a name and IP — this creates a manual target with all three checks enabled.
2. Click the device in the sidebar. A tab opens. If it is not monitored yet, click Enable Monitoring.
3. Ping starts immediately. For bandwidth, configure SNMP credentials — see [[crearack—monitoring—configurar-snmp]]. For web checks, set the URL in HTTP Settings.
4. Optionally define thresholds via the Alerts buttons — see [[crearack—monitoring—alertas]].

Within minutes the charts fill with data and you can browse history — see [[crearack—monitoring—dashboards]] and [[crearack—monitoring—metricas]].

Taking a device out of service

Some equipment is only used part of the time — reinforcement access points that go out for an event and come back to the storeroom, a spare UPS, a screen that is packed away between shows. A device that is switched off on purpose is not a fault, and it should not look like one.

Open the device and click Take out of Service, then confirm. From that moment the device:

• stays in the list, greyed out — it does not disappear, and it keeps its card and all its history;
• raises no alerts, notifications or CNS insights, and is not counted as down: it drops out of the offline count, the health score and the “worst” rankings. Fleet reports (Wireless, UPS, Signage) list it in grey as Out of service and count it on its own card, and the Help assistant names it apart instead of reporting it as down;
• is only pinged, lightly, so CreaRack can tell when it is back.

You do not have to remember to bring it back. As soon as the device answers ping again it returns to service by itself and is monitored as before. You can also click Return to Service at any time.

Two things worth knowing: if you take a device out of service while it is still switched on, it will not bounce straight back — it returns only after it has been seen offline and then answers again. And a device that blocks ping (ICMP) cannot announce its own return: bring it back with Return to Service. Devices out of service still count towards your plan.

Troubleshooting

• Device missing from the sidebar — it has no IP or no network management in CreaRack. Use Add to monitor it as a manual target.
• Charts stay empty — checks have just started; data appears after the first refresh cycle. For SNMP, verify credentials first.
• Monitoring stops when you close the browser — that is Cloud Only behavior by design. Switch to Sentinel mode for 24/7 coverage.

Related

• [[crearack—monitoring—dashboards]] — Overview dashboard and device detail view
• [[crearack—monitoring—metricas]] — collected metrics, intervals and retention
• [[crearack—monitoring—configurar-snmp]] — SNMP polling setup
• [[crearack—monitoring—alertas]] — alert thresholds and events
• [[crearack—monitoring—ping-http]] — ping and HTTP checks
• [[crearack—monitoring—cns-sentinel]] — the AI analysis layer
• [[crearack—monitoring—itsm]] — incident management
• [[crearack—monitoring—fleet-manager]] — Local Agent fleet
• [[crearack—monitoring—deep-discovery]] — vendor-aware deep discovery
• [[crearack—network—device-page]] — the device card and its polling cadence
• [[crearack—monitoring—que-es-wireless]] — Wireless Monitor (own module)
• [[crearack—monitoring—que-es-ups]] — UPS Monitor (own module)

Véase también

• [[crearack—monitoring—dashboards]]
• [[crearack—monitoring—metricas]]
• [[crearack—monitoring—configurar-snmp]]
• [[crearack—monitoring—alertas]]
• [[crearack—monitoring—ping-http]]
• [[crearack—monitoring—cns-sentinel]]
• [[crearack—monitoring—itsm]]
• [[crearack—monitoring—fleet-manager]]
• [[crearack—monitoring—deep-discovery]]
• [[crearack—monitoring—que-es-wireless]]
• [[crearack—monitoring—que-es-ups]]